Google Gemini's Image Generation Gets Major Upgrade with Imagen 3 and Resizing Options

Gemini image generation just tipped for a major upgrade -- here's what's coming

Code found in a recent beta hints that Google is working to fix one of the most annoying parts of using Gemini's Imagen 3 image generation. Google Gemini only recently got the ability to create images, and the results are pretty impressive. However, one continuing complaint about Gemini is that it only generates square images. Currently, Gemini's images are all formatted in a 1:1 ratio, and the only way to change this is to crop the image yourself. However, Android Authority has found code lines that hint this may be changing. In a recent report, there were lines of code discovered in the 15.41.34.29.arm64 beta of the Google app that reveals some in-development settings, including being able to select from several different sizes. According to the report, we'll be able to resize the image to the most common ratio options, including 16:9. It's worth noting that these settings don't simply crop the original image, instead, Gemini goes back to the original image and adds to it to fill the new space. You can even resize the image again and keep the same expanded environment, just with a shrunken picture subject. One thing to note is that the current Imagen model is free with Google Gemini, but the basic version can't create humans, portraits or faces. If you want to generate those kinds of images then you will need to invest in Gemini Advanced, which you can try for one month free. An added bonus is that Gemini saves its images as a jpg, unlike ChatGPT. There's been a fair bit of news regarding Gemini recently, one of the biggest was the release of Gemini Live to all compatible Android devices. Gemini Live stands out from other AI by being able to hold continued conversations with users, which helps to make it feel more natural. Added to this is the new look for the Gemini app which has streamlined the user experience tremendously. This change will be one that will make a lot of users happy, and improve the experience of using Gemini on one of the best Android phones.

Google is finally fixing one of the most frustrating things about Gemini's AI image generation

Key Takeaways The latest beta version of the Google app offers new Gemini image resizing options. There are major ethical concerns over AI usage in creative sectors, especially regarding copyright issues. Google Gemini is on almost every single Google Workspace app. Google Gemini is one of the biggest new players, if not the biggest newcomer, to the artificial intelligence landscape of 2024. It's Google's attempt to grow its multi-industry-leading company in another realm, and it's doing so by bringing Gemini to practically every app in the Google Workspace. While AI is certainly cool, it represents a horrible reality for many in creative sectors, as the ethical questions over how AI is trained crosses over with both copyright issues and fair pay disputes. Imagen 3, Google's text-to-image generator, is directly involved in this debate, but Google said it will defend you if its AI images infringe on your copyright. Never mind the ongoing debate; Gemini's image creation user experience is getting better, and it now offers more customization options to its users. Related ARTICLE: Google updates Gemini with the Pixel Studio app's image generation model Imagen 3 is here 1 For as long as you can generate images using Gemini, the common denominator between Imagen 3 and Imagen 2 has been its square canvas outputs. That's set to change in an upcoming stable version of the Google app, as Android Authority reports that, on version 15.41.34.29.arm64 beta, there are now in-app settings for resizing Gemini images. It doesn't look like a simple crop and resize tool, either; the site reports that Gemini will add more background detail when cropping down and fill in the blank spaces when making images larger. Gemini's busy year of image generation (Source: Android Authority) There's a lot going on behind the scenes at Google with Gemini's image creation tools, and Imagen 3 is getting smarter every day. The third version of Imagen was released in August, and it improved upon Imagen 2's prompt comprehension and added a four-image output option called Imagen 3 Fast. One shortcoming (or, something that we should be happy about as a society) is that Gemini cannot generate images of people. Google admitted that Gemini's image generation capabilities "missed the mark" early on, and while images of people still cannot be generated, we think that's A-OK. Imagen 3 has seeped into many different Google apps, like Slides. Users don't have to leave the confines of their template to make AI images in the cloud, but they do need to have an Enterprise or Education account, or access to Google Workspace Labs. Personal accounts can gain access to the feature by subscribing to Google One AI Premium. Other than Slides, the Google Pixel 9's Screenshots app finally got the Gemini integration it should've had all along.

Google Gemini just got AI image generation back with Imagen 3 -- how to try it now

All Google Gemini users can make images using Google's latest artificial intelligence image mode, Imagen 3. This includes those using it on the web, in the app or integrated into Android. It was previously only available to Gemini Advanced subscribers using the web version of the AI. Imagen 3 was unveiled by Google during the I/O event earlier this year. Built by the company's research lab, DeepMind, it can create square format images with a similar level of realism as those from Midjourney or Flux. There are some limitations, especially for free Gemini users. At the moment, you won't be able to use it to generate images of people unless you pay $19 per month for Gemini Advanced, and even then, it won't make images of real people. Gemini's true power in terms of image generation comes from its ability to easily adapt or even completely change an image from a simple text description. You may need to update the app to get image creation abilities. Imagen 3 is a visual upgrade on the previous Imagen 2. The images are richer and more detailed, and the model is better at following instructions given to it by Gemini. Google removed image generation capabilities from Gemini for some time over concerns it was being overly cautious when rendering pictures of people. The previous model also tended to make everything look like it was plastic. The new model brings Google's image generation capabilities in line with DALL-E 3 from OpenAI, although it still only generates square images, whereas ChatGPT can use DALL-E 3 to make pictures of any orientation and size. In my experiments, I found it still refused more often than I'd like to make an image, especially if it was something more complex, such as showing a person sitting in a cafe on a moon base. But for everyday depictions -- it is great. Adding image creation comes off the back of another update, making it easier to send an image to Gemini for analysis. This is known as AI Vision and allows the model to describe the contents of a picture. It can be beneficial for spotting a missing object in a room or creating a recipe from ingredients.

Gemini will soon let you create images in the shapes and sizes you want (APK teardown)

Users can also resize Gemini's output, allowing the subject to take up less of the frame. Gemini is Google's attempt at bringing powerful, modern AI to the masses, and just as just as you'd expect from a robust generative model, it's pretty handy at dreaming up images. All you have to do is ask Gemini to make a picture for you, and it'll cook something up in just a few moments. With the upgrade to Imagen 3, Gemini's output is looking better than ever, and now we're checking out one way the app could soon give you even greater control over the kind of pictures it generates.

Google updates Gemini with the Pixel Studio app's image generation model

Key Takeaways Initial AI applications like chatbots and image generation helped popularize apps like ChatGPT and Midjourney. Google now offers deep AI integrations in its apps and Gemini chatbot assistant for image generation, besides other things. Imagen 3 on Gemini now provides free access to realistic text-to-image models, but may seeem like a disservice to recent Pixel buyers. AI continues evolving but just a handful of initial applications captivated the masses -- chatbot-style interactions and image generation from textual prompts. These factors alone helped the likes of ChatGPT and Midjourney skyrocket in popularity. Now, Google has several deep AI integrations in its apps, as well as a chatbot assistant called Gemini that can handle image generation too, making it one of our favorite AI apps for Android. The company recently gave Gemini's image generation abilities a shot in the arm. Related Pixel Studio is ready to work its AI magic on images from apps you actually use Easier ways to be mindless Google pulled the wraps off Imagen 3 at I/O earlier this year, and the tool has since been used behind the scenes for apps like Pixel Studio. It allows users to prompt the AI to generate images, and then selectively refine the result in multiple iterations. The app seems to work well with images of people, but was limited to the latest Pixel 9 series. A recent visit to the Gemini service on the web revealed the service also gives access to Imagen 3, even on the free tier (via NotebookCheck). Google had been testing Imagen 3 on Gemini for the last few months, through the AI Test Kitchen. After a few months of testing, the company seems to have decided to give all users access to these image generation capabilities, because we are seeing a pop-up about it on Gemini. While the Pixel Studio app converted stylistic definitions of your prompt into drop-down menus with alternatives, Google's deployment on Gemini instead focuses on image sharpness, detail, and more photorealism than the outgoing model. Fast progress might seem like a disservice to paying customers On the one hand, free access to Google's flagship text-to-image model with surprising realism is a huge plus, there's a non-zero number of customers on the flipside who bought the Pixel 9 series to enjoy the convenience of the Pixel Studio app. The phone isn't even a year old at this point, and Google making Imagen 3 so easily accessible might seem like a disservice to the Pixel buyers who forked real cash for the AI smarts. However, given the rapid rate of advancements in AI, we aren't surprised Google is happy to offer Imagen 3 to the masses for free while it probably works on a different generative model with better capabilities for the next generation. In such cases, there's a strong argument to call such features 'bloatware' on phones if they are available for free just months later. That said, Imagen 3 on Gemini is a step forward for people using the chatbot to whip up images for a college presentation or just memes to keep the group chat lively.

You Can Now Generate Images on Gemini Using the Imagen 3 AI Model

Google said Imagen 3 has improved prompt adherence and text rendering Google announced a significant upgrade for Gemini, its in-house artificial intelligence (AI) model, on Wednesday. The company announced that the image generation capability of the chatbot will now be handled by the Imagen 3 AI model for all users. Imagen 3 is the Mountain View-based tech giant's latest and most capable image generation model. Apart from the Gemini app, the feature is also being extended to the API version of Gemini to let developers build apps and experiences based on this capability. In a post on X (formerly known as Twitter), the official handle of the Google Gemini App revealed that all users, including those on the free tier, will be able to generate images using Imagen 3. The post highlighted that the AI model offers a high degree of photorealism, better prompt adherence, and adds fewer unwanted elements to images. Gadgets 360 staff members were able to verify that the Gemini app is indeed using Imagen 3 to generate images. To test its capabilities and compare it with Meta AI, we gave both chatbots the same prompt. The prompt was, "Draw an image of a golden retriever dog sitting on a train berth, looking out through the window at the Alps. The train has a wooden interior and the seats are green in colour. All other passengers on the train are also animals. One human conductor is checking for tickets." The generated images can be seen above. While both AI models failed to incorporate one or more elements instructed in the prompt, Gemini was able to incorporate more elements. Additionally, while Meta AI generates images in 1280 x 1280 resolution, Imagen 3 images are generated in 2048 x 2048 resolution. Imagen 3 can generate images in a wide range of styles such as photorealistic, textured oil paintings, and claymation scenes. Users can also request images to appear as if it has been taken from a specific camera such as a Nikon DSLR, GoPro style, wide-angle lens, and more. Google has said that the AI model comes with inbuilt safeguards to reduce the risk of deepfakes. Every generated image also comes watermarked with SynthID, a technology that adds an invisible AI label within the pixels of the image. It cannot be cropped out or removed and is present even in screenshots.

Google rolls out Imagen 3 for all Gemini users - Phandroid

One of the standout features of AI is its ability to generate images using nothing but word prompts. If you're looking to try some image generation using AI, then you might be interested to learn that Google has rolled out Imagen 3 for all Gemini users. For those unfamiliar, Imagen 3 is an AI image generation tool by Google. The updated model is said to offer a higher degree of photorealism, fewer distracting artifacts, and what Google claims are "better instruction following". Also, it comes with text rendering, which Google says will give users a way to create more personalized images with AI, like greeting cards, and so on. Imagen 3 will be available for all Gemini users, even if you are using the free version of the AI. However, there are some limitations. For instance, you won't be able to generate images of people if you're using the free version of Gemini. You'll have to upgrade to Gemini Advanced. It will cost you a subscription fee, but in addition to generating images of people, Gemini Advanced offers additional features like priority access to new features, 2TB of cloud storage, run and edit Python code in Gemini, and more. If you're interested in trying out Imagen 3 yourself, just launch Gemini on your phone or computer and use prompts like "draw", "create", or "generate" along with a description of whatever it is you want the AI to draw.

Google Gemini's new AI image generator just rolled out to everyone for free - with one annoying limitation

Google's AI image generator Imagen 3 is now available to all Gemini users on mobile or desktop, for free. So, if you ask Gemini to create an image for you it will now use Imagen 3 by default, even on the free tier, but there's one annoying limitation. Google calls Imagen 3 its highest-quality image generation model yet and claims it's better at following your instructions and produces fewer visual artifacts than its previous models. However, there is one annoying limitation - if you want to generate images of people you need to be a Gemini Advanced customer. Gemini Advanced costs $19.99 (£18.99, AU$32.99) a month and comes as part of the Google One AI Premium Plan that includes 2TB of storage and Gemini in Google Docs. Unlike some other AI image generators, Imagen works by generating just one image from each prompt. You can refine the image by requesting changes, so it remembers what you asked for previously, but a new image is generated for every prompt you enter. If you like the preview you see you can download the image and you'll get a 2048 x 2048 resolution JPEG. You can download as many images as you like because there's no specific daily limit for image creation, even on the free tier. The main restriction, of course, is that you can't create images of people. Animals however aren't a problem. Of course, once you upgrade to Gemini Advanced you can create as many images of people as you like. Imagen 3 has a range of features that are worth exploring. For example, you can ask it to create photorealistic landscapes, richly textured oil paintings, or even claymation scenes. Google is very proud of Imagen 3's great text rendering capabilities. So, if you ask it to create a comic book panel with a speech bubble coming from one of the characters, the text should be readable. However, we found that it still had some work to do in this area. For example, the prompt "create an image of a panel from a 1950s sci-fi comic where a man and woman look out at a futuristic city and a speech bubble from the man says "I never thought Mars would have such a great atmosphere", generated this unfortunate image: You can also ask it to create words made out of other shapes, or carved into buildings and it should look realistic. All these improved features mean that Imagen 3 is going to deserve a place amongst the best AI image generators.

I just created 5 images with Google Gemini -- and it left me both impressed and annoyed

Google's Imagen 3 has finally arrived in Gemini and is already making waves with its ability to create stunning visuals based on simple prompts. Google boasts that it's their "highest quality image generation model yet," and I couldn't wait to explore its full potential. AI-driven image generators dominate the AI landscape, with models like DALL-E 3 in ChatGPT and Midjourney getting most of the attention. Gemini previously had access to Imagen 2 but this was removed after some issues with performance. The new model, built by Google's AI lab DeepMind, offers a fresh take on the process with its sophisticated approach to visual creativity. My goal was to see how well it handles diverse prompts, from the texture of sushi to the intricate features of a human face. I was impressed by its realistic images but also encountered some quirks that reminded me this is still a developing technology. Here's what stood out. Using Google Gemini's Imagen 3 is straightforward. The interface is intuitive, and I like that it allows for prompt adjustments or re-generations so if the initial result isn't perfect, it's easy to iterate without starting from scratch. This balance between speed and flexibility encourages creativity and exploration. I decided to experiment with a mix of both detailed and open-ended prompts. I figured that this was the best way to test Gemini's creativity while determining if it responded better with vague or more explicit prompts. Interestingly, I was impressed by the visuals in some cases while noticing an obvious lack of nuance at other times. Prompt: "Create an image of a plate of sushi." I started with something simple -- a plate of sushi. Gemini created a beautifully detailed image, with vibrant colors and textures that made the sushi look appetizing. I was incredibly impressed by the detail and how the sushi looked as if it was taken from a print in a magazine. However, it still lacked the creativity I was hoping for because the image seemed quite generic. Yes, the realism was impressive, but it didn't push boundaries or showcase much variety in artistic style like I had hoped. Prompt: "Create an image of a cozy living room." I decided to be a little vaguer with this one to see what Gemini would create. The result was hilariously bad. The classic furniture, harsh lighting, bland color schemes, and oh, yeah, the chandelier did not set the mood I was hoping to achieve. When I think of "cozy" I want a big comfy couch, big windows, soft lighting and a warm blanket. This image was completely off, but was still impressed by the details, placement of the objects, and overall aesthetic of the room -- despite it not being the room I wanted. Prompt: "Create an image of a majestic tiger in the wild." For an animal image, I requested a majestic tiger in the wild. The AI delivered with the fur details but from what I could tell, there was not natural setting. It simply looked like a tiger taking a school picture. Don't get me wrong, it's a visually striking image. However, the tiger's face had a slightly unnatural look, which distracted from the otherwise impressive composition. Where the AI excels at in textures, it lacks in emotional expression. Prompt: "Create a retro-style movie poster" In my opinion, this image is where Imagen really shined. It nailed the assignment with a truly eye-catching design. It's so good that I would wear this design if it were on a t-shirt. The bold font with vibrant colors that faded into the edges created an authentic retro vibe. The creativity here was simple, yet it seemed as though Gemini grasps the nuances of poster design. So far, this was the most satisfying of all the images I generated. Prompt: "Create an image of NYC" Lastly, I wanted to see how it would do with architecture, so I went with NYC. Gemini generated a hyper-realistic image that could have been a photograph. The image was a technically incredible visual of the skyline, but the sky itself looked almost too perfect, lacking the imperfections that might make it feel less AI-generated. While the image was impressive, it didn't quite cross the line into believable realism. Overall, Gemini's Imagen impressed me. The detail, texture, and design aesthetics were pretty amazing. However, it's clear that it needs extra prompting to truly nail down the final visual. It struggles with more nuanced requests and can't seem to creating unique, artistic interpretations. I like that Imagen is free, but keep in mind that the basic version cannot yet create humans, portraits, faces. If you're looking for that feature, you'll have to go with Gemini Advanced, which you can try for one month at no charge. As a bonus, unlike ChatGPT, the images are saved as jpg. Imagen shows a lot of promise, and with continued updates, it could become a powerful creative tool. However, for now, some aspects feel a bit robotic and slightly off, leaving room for improvement.

Everyone everywhere can now enjoy Imagen 3 in Gemini apps

Google also claims Imagen 3 reduces distracting artifacts, which are usually dead giveaways that a picture has been generated with AI. However, Google is also cognizant of the dangers of creating more accurate images of humans. We have discovered evidence that Google is working to bring AI image detection right into Google Photos, for example. This should, hopefully, make it easier for people to easily tell if an image depicts something that really happened or is just a visual creation of AI.